Qualcomm Execs Discuss Plans For Growth In Automotive, IoT And Beyond

Qualcomm has spent 40 years pioneering new types of mobile chip technology, and it is intent on maintaining its leadership there. But can it translate that dominance into similar success — and scale — in other markets?

NurPhoto via Getty Images

Qualcomm celebrated its 40th anniversary this year. Across those four decades, the company built its reputation as a supplier of chips for mobile phones. But while it retains a dominant position in that market, these days an increasing amount of its business comes from other rapidly growing sectors, including automotive, IoT, PCs, the datacenter and, on the horizon, robotics.

Yet even after supplying all the processors for the first wave of Copilot+ PCs a year ago, and after its recent deal to equip datacenters in Saudi Arabia, and even with a well-established automotive business that grew 55% last year and has a $50 billion pipeline of orders, Wall Street seems unconvinced by Qualcomm’s diversification story. So while its leaders received the honor of ringing the bell at the stock market on its anniversary, their work hasn’t been similarly rewarded in terms of the company’s valuation. What gives?

In recent weeks, I’ve had the opportunity to interview four of Qualcomm’s top operating executives — each of them with more than 20 years at the company — in part to talk about the defining characteristics that have set Qualcomm apart so far, but with a greater emphasis on where the company is headed across multiple new markets in this era of AI. These interviews offered extraordinary access to the company’s strategic thinking as it tries to extend its ecosystem-based approach into one new sector after another.

Note that Qualcomm is an advisory client of my company, Moor Insights & Strategy. That said, these interviews were conducted on the record, and Qualcomm was consulted during the editorial process only to confirm the accuracy of the technical details and quotes provided here.

Qualcomm’s Origins As A Wireless Innovator And ‘Ecosystem Company’

All four of the executives I talked to have been with Qualcomm since the 1990s or early 2000s, and all of them described it as ecosystem-oriented from the start. Nakul Duggal, who runs Qualcomm’s fast-growing automotive and IoT businesses, just celebrated his 30th anniversary there; he told me that “I’ve always seen Qualcomm as an ecosystem company.” Durga Malladi, who oversees technology planning and coordinates the company’s AI initiatives — besides running its new datacenter business — joined Qualcomm after earning his Ph.D. at UCLA in 1998. He summarized the engineering mindset back then as, “We know we are a systems company, and so we’ve got to think from an overall systems perspective.”

CFO and COO Akash Palkhiwala, who joined in 2001, framed Qualcomm’s approach in contrast to many other tech companies. “We are a creator of ecosystems; we are a creator of technologies,” he said. “Our role has always been about creating ecosystems, about creating technology — and enabling the industry with those things.” Even though I interviewed each of the executives separately, they all came back to this theme without prompting again and again, applying it just as much to the company’s current areas of expansion as its past successes.

Alex Katouzian, who joined the company in 2002 and now runs its Mobile, Compute and XR Business Unit, connected this theme back to the company’s early days in wireless. As an undergraduate at the University of California, San Diego years before, he had studied under Qualcomm cofounder Andrew Viterbi, who helped develop the CDMA standard. Viterbi and other engineering leaders at the company first “realized that CDMA was a better air interface,” then “also realized that this air interface is good for data” — which, in time, led to the wireless internet. Katouzian credits these as “massive, massive pivotal points” in the technical foundation of the company.

Crucially, the leaders in those days didn’t stop there, and Katouzian explains how their next steps set the dominoes in motion for Qualcomm’s prominence in the wireless market — and its way of doing business even today. “After proving out CDMA, what they did was they built the entire thing,” he told me. “They went vertical. They built chipsets, they built the infrastructure. They built the phones, they had the software. They had the whole thing.” Realizing how chipsets, software and IP licensing could work together to cement the company’s commercial position, they concentrated on those things, which led to embedding technology and multimedia within the chipsets — and hence, Katouzian said, the capability of the phone to become a camera, a gaming machine, a video capture device and an interface to social media.

Later, this led to the birth of the company’s Snapdragon brand, as well as its movement into other end uses. As he put it, “Once I have multimedia, graphics, camera, gesture, video capture, then I can go into adjacent markets. And then that led to auto[motive], IoT, XR, PC and all the adjacent markets that we’re trying to get into.” Time and again, each of the executives I talked with came back to how Qualcomm’s long history in designing for ecosystems that require high connectivity, small form factors and tight power management has set it up for success.

Qualcomm’s Present And Future In Smartphones, Wearables And PCs

Today, Katouzian oversees the part of the business that serves makers of smartphones, other mobile and smart devices (including XR), gaming gear and PCs — and all the Snapdragon platforms. Thinking about new types of mobile experiences, he believes that the smartphone will remain central as a “very big computing device that’s with you all the time,” unlike your car or your PC. This makes the smartphone the natural point of connection for smart glasses, watches, earbuds and so on.

Increasingly, these devices will work together to provide users with cool new experiences like shooting video with your smart glasses and sharing it to social media through your phone. At some point, it will be routine to do things like this using voice or gesture commands, and without even touching the phone in your pocket.

Beyond these new types of UX, though, AI assistants will increasingly operate across a user’s devices to create highly personalized experiences and bodies of data. If you follow advances in AI models, it’s not hard to imagine the inferencing specificity of a model informed by the data coming from a set of smart glasses that sees and hears everything you do, plus a smartphone that knows pretty much everything about what you do in a day. While there are obviously important safeguards to have in place (Katouzian is also responsible for Qualcomm’s security technology), this will surely be one of the ultimate expressions of the company’s ecosystem mentality.

Creating cross-device experiences is an area of acute focus for OEMs trying to compete with Apple when it comes to the stickiness of their products. I’ve seen this in action for years, not least in my own use of Samsung Galaxy laptops, phones and watches, but Katouzian hit me with a startling statistic. “If you buy three devices from the same manufacturer,” he told me, “the likelihood of you sticking with that manufacturer is over 95%.” Apple nailed this years ago with easy syncing across iPhones, Apple Watches, AirPods and so on. Qualcomm is now aiding OEMs in the Android universe, including Samsung and a slew of Chinese companies like Oppo, Vivo and Xiaomi, by building end-to-end reference models — complete with Qualcomm’s cutting-edge perception technologies — to help them nail down specific interactions and use cases across these devices.

We also talked about how PCs come into this picture. There’s not enough space here to go into the details of Qualcomm’s breakthrough role as the sole provider of processors for the first wave of Copilot+ PCs that launched a year ago. (I’ve written about that at length elsewhere, and I’ll touch on it a little more in the following section about AI.) But during our conversation Katouzian made the point that AI PCs will also sync with smart wearables to enable new use cases. For instance, a technician performing field repairs might look at a piece of broken equipment using smart glasses; the glasses could then use the technician’s PC to connect to the cloud for specialized AI models and technical information. Imagine how much easier the task will be when the smart glasses visually overlay a troubleshooting schematic on top of the faulty equipment, helping the technician pinpoint the trouble and make a quick repair.

Use cases like these, and the AI PCs that can handle them, are not way off in the distance. And while the first generation of Copilot+ PCs already began to open up on-device AI functionality — and revolutionized PC battery life for road warriors like me — Katouzian is bullish about what we will see when second-generation Snapdragon X Elite-powered PCs come out in September.

How AI Fits Into Qualcomm’s Strategy

These days, Qualcomm focuses on being an AI-led company in everything it does. Durga Malladi spends a lot of his time thinking about which parts of AI need to happen locally and which parts should happen in the cloud. His fundamental view, and Qualcomm’s, is that AI must be hybrid. “One of the earliest viewpoints that we had was for AI to scale — we reached a conclusion that it cannot be just in the cloud,” he told me. “It has to have a piece of the puzzle in the devices as well, and not just because [that] serves us. I mean, it’s a nice narrative, but that’s not the reason why.”

He explained that, as consumers start using AI in all of their devices, they’re not thinking of it as AI, but as a set of functionality that “just happens to be there all the time.” They come to rely on it, and don’t want their user experience to be defined by the quality of their connection to the cloud at that moment.

If you’re in a place with great connectivity — Malladi gestured to the corporate meeting room he was sitting in — you might not need to run anything on-device, because some enterprise server will handle it. But what if you’re using your laptop on a transatlantic flight? “Regardless of the connectivity, we’ve got to be able to make sure that AI experience remains,” he said. “So that means you have to have the right kind of processing in all the devices.”

He also made some acute observations about how quickly AI models are shrinking to deliver the same performance. “You know, DeepSeek from China made such a big deal back in January,” he said. “That was a 671-billion-parameter model, and I remember someone asking me, ‘So what does this mean? Does it mean that in the U.S. we are losing our leadership?’ I was like, ‘Wait for two months.’ And actually, it only took two months before a 32-billion-parameter model was already outperforming that 671-billion-parameter model.”

Obviously, this progression tends to favor the idea of increasing the use of powerful smaller models on personal devices over time. More to the point, AI will become so ubiquitous that it will be fundamental to the experience of using those devices. In this context, Malladi said that Qualcomm coined the phrase “AI is the new UI.” He noted that, besides having a nice ring to it, “It’s a true statement.” He pointed out that the latest big change in UI was the rise of touch interfaces on smartphones 20 years ago. Before that, it was the move from green screens to GUIs 20 years before that. So he believes we’re due for another major UI shift.

Diversification Principles — And Success In Automotive

Automotive, smart wearables and even AI on personal and industrial devices can be seen as direct extensions of Qualcomm’s original wireless business and ecosystem approach. But what is the thinking that governs its diversification beyond mobile in recent years and today? For COO Palkhiwala, it comes down to four principles:

1. Continue to lead — or “set the design point” — in handset technology. (I would add this corollary: Enter only those markets where you can capitalize on what you’re already good at.)
2. Identify and target industries where there is incredibly high leverage of technology.
3. Target only those industries that are going through an inflection point.
4. Build the go-to-market muscle that is required for that industry.

During my interviews as well as my other interactions with Qualcomm’s leaders over the years, I can attest that they are highly disciplined about applying these principles. This leads again and again to that moment of recognition when you look into what they’re proposing in a new market and say, “Oh, right, that does fit.”

The easiest example of this is the automotive industry. Qualcomm’s automotive chief, Nakul Duggal, explained the importance of first understanding a new market thoroughly. “You actually have to understand what ecosystem you are a part of and what the needs of that ecosystem are,” he said. “So we took that approach with automotive almost a dozen years ago, and we spent essentially all of our time … just understanding how this industry works, the people, the tier-ones [suppliers], the OEMs, the tier-twos, the software players. We did it globally.” This approach helped them understand how to build a GTM process for the automotive business while they also built relationships “at a time when nobody [in our space] was really spending a lot of time on auto as an industry.”

That industry was certainly facing an inflection point — and entering a period where the leverage of technology was only increasing. I’ll let Duggal explain it: “What became pretty apparent was that this was an industry that was probably five years behind the rest of the [tech] ecosystem in terms of adoption of new technology, partly because it had an incumbent ecosystem that was fairly traditional, partly because it was risk-averse, partly because it hadn’t really seen any major catalyst for destruction. And then as Tesla started to become very relevant in the U.S. and EVs started to become very relevant, there was certainly enough of a catalyst for people to start to think outside of the box.”

That covers principles 2, 3 and 4, but what about capitalizing on what the company is already good at? Looking back, Duggal believes that Qualcomm positioned itself well “in terms of shaping the company’s pivot from a consumer company that does smartphones, to a systems company that actually does all types of silicon.”

He sees it as a multi-step journey, starting with connectivity. “We took our smartphone roadmap and we converted it to a digital cockpit roadmap and brought an ecosystem along with it.” Then, however, they decided they needed to expand the focus to include driver assistance, safety and (ecosystem alert!) “building the stack.” Duggal believes that Qualcomm’s journey “has allowed us to embrace full diversification towards a new ecosystem, while at the same time leveraging a lot of the core capabilities and competency within the company.”

In the decade-plus since then, the company has continued to expand its automotive capabilities through a mix of organic and inorganic growth. Palkhiwala used the automotive example to explain Qualcomm’s build-versus-buy logic. “We built chipsets organically because we already had the core technology,” he said. “We built software for digital cockpit organically, because we are big experts at Android, and we know how to do that. But where we did not have the talent was to build a stack for ADAS. And so in that case, we acquired competence from someone who had been in the industry and knew how to scale that market.” Speed is a major consideration: “If we see a missing piece of technology that has an established industry technology, we don’t want to wait and develop it organically.”

Speed is especially important given the rise of the Chinese automotive industry, which moves fast. Just like Tesla was disrupting the U.S. auto market a decade ago, BYD and others are doing the same to the global auto market today. I’ll cover this in more detail in an upcoming deep dive, because Duggal and I discussed this for a while. The thing to note for now is that Qualcomm is acutely aware of the challenges of building stacks that accommodate all the regional variations in markets, AI use and even physical geographies — and it’s doing something about it.

“We went to China in 2023 and we realized that we were a little bit behind on ADAS,” Duggal told me. “Nobody was using our ADAS silicon digital after Covid [then], but we are now deployed across 20 different Chinese automakers.” Qualcomm also has a big partnership with BMW that is bearing fruit. The results? “We’re going to be in 60 countries by the end of the year, in 100 countries next year.”

Expansion Into IoT, The Datacenter And Robotics

Duggal understandably has a reputation for getting things done, so he’s now added IoT products to his portfolio. The company’s definition certainly includes the kind of smart home and industrial applications that we typically think of for “IoT,” but it is broader than that. The five areas that Duggal and his team have defined are industrial connectivity, consumer and commercial, cameras, industrial processors and robotics. (More on robotics in a moment.) They have also targeted a number of specific verticals including retail, industrial automation, public safety and energy and utilities.

Again, it’s not hard to see how these specific areas relate to the four principles of diversification that Palkhiwala laid out — especially as new IoT paradigms incorporating AI take hold and potentially disrupt many different sectors. Qualcomm’s expertise in delivering high performance at a specific power level also seems tailor-made for IoT. To put it another way, whereas plenty of chip makers are constrained in what they can offer at the network edge and on individual devices, Qualcomm has been thriving in that zone for decades. And again the team has gone out of its way to understand the markets being entered, with different GTM actions for “certain customers that know what they’re doing” and need less support, as Duggal put it, versus others that “are not as conversant” and need Qualcomm to build out the stack rather than only providing silicon.

If you’d asked me a few years ago, I would not have been as optimistic about Qualcomm’s recent entry into the datacenter market. When we talked, Malladi pointed out in good humor that the company has tried this more than once in the past. Yet its recent deal with the Saudi-backed AI company Humain — announced in connection with President Trump’s visit to Arabia — suggests significantly more traction this time.

I don’t expect Qualcomm to overtake Nvidia or AMD (the two other chip companies that announced big deals on the same trip) in the datacenter anytime soon, but it does bring something special to the table because of its long history of lowering energy consumption in virtually everything it does. “We feel like that’s a new angle that we bring to the datacenter for everything that the incumbents today have,” Malladi explained. “We say, ‘I’ll match your performance and I’ll lower the energy consumption.’ That’s where things are different for us.” I can see that message resonating louder and louder as AI datacenters continue to gulp down energy at rates beyond what we’ve ever seen.

Like the datacenter, robotics represents a potentially huge market for Qualcomm. Palkhiwala thinks it’s a natural fit for the company’s diversification principles. The robotics industry is undergoing a massive transition from task-specific, non-mobile robots — think of the big ones that install body panels on automotive assembly lines — to autonomous general-purpose (and often humanoid) robots. The latter will rely on low-power, high-performance CPUs, GPUs and wireless technology to employ AI at the edge. More than that, they will need good cameras, not to mention elements of ADAS and XR. So even in this simple description, we’ve touched on about ten different areas where Qualcomm already has world-class expertise. Plus it’s a market driven by highly technical OEMs, which again sits right in Qualcomm’s wheelhouse in terms of GTM.

Let me emphasize that it’s early, early days for robotics of the type we’re talking about. But I also don’t think it’s an accident that the two big chip companies talking about robotics the most this year are Nvidia and Qualcomm — and I won’t be surprised if the two of them make hay in the robotics market in years to come.

How Do You Value A Company Like Qualcomm?

Mark my words: history will show that getting compute as close to the data as you can is the most efficient thing you can do. Irwin Jacobs, Andrew Viterbi and their colleagues may not have thought of it that way when they were launching Qualcomm 40 years ago, but the company has positioned itself very well indeed to take advantage of this truth, which I believe is even more relevant in this era of — hybrid, high-speed, always-on — AI.

Qualcomm’s success in mobile means that a lot of people think of it as the mobile chip company — and it is. But as we’ve seen, it has also been very thoughtfully diversifying for many years, with growing success. (For more numbers on that, see my writeup of its investor day in late 2024.) And its position in the technology landscape is unique. During our conversation, Malladi said, “We already are there in products that can range from doorbells to datacenters. Now think of one company in the industry that has that kind of scale or the ability to bring in those kinds of skills.” The fact is, Qualcomm is alone in that position, because there’s nobody else who does that.

This is hardly a matter of touting uniqueness for its own sake, because the company keeps showing how practical its differentiation is across mobile devices, PCs, automotive, IoT and so on. Malladi put it very plainly: “In effect, we have invested significantly in energy-efficient high-performance computing. It’s not just high performance by itself, but energy-efficient high-performance computing.”

That combination of efficiency and performance speaks for itself in mobile devices. But it has also completely changed the battery-use paradigm for anyone using a Copilot+ PC. It will most certainly matter — even up to a geopolitical level — when it comes to AI datacenters. And Qualcomm’s engineers have been designing on this basis week in and week out since the 1980s.

The operational challenge now is for Qualcomm to maintain its level of technical and business performance as it scales up. I’m talking about scaling numbers of chip cores from tiny mobile applications, through the intermediate levels of PCs, software-defined vehicles and edge servers, all the way up to the big iron of the datacenter. And I’m talking about scaling up in terms of the number of industries it serves at volume. I like the company’s chances, considering the track records of these executives, their boss, Cristiano Amon, and their predecessors going back to Jacobs.

When will that be enough for Qualcomm to achieve the kind of success on Wall Street already enjoyed by technology darlings like Nvidia? I don’t know. Maybe the company is taken for granted, or maybe its non-mobile businesses don’t yet represent large enough chunks of its top line. Maybe the uniqueness of its position in the landscape — and why it’s important — isn’t clear enough. It’s not as splashy as Tesla in SDVs, as famous as Intel or AMD in CPUs or as lava-hot as Nvidia in AI.

Palkhiwala — who, you’ll recall, is CFO as well as COO — seemed philosophical when I asked him about it. “I think investors largely understand our diversification strategy, but they want to see it play out in numbers,” he said. “And so we are heads-down, focused on what we told investors we will do.” He believes that continued execution will provide the Street with the proof points to support a higher valuation.

“The technology portfolio [we have] is ‘one of one’ in this industry,” he added, “and that is what matters in the long term — really being able to leverage that portfolio to execute on the plan we’ve already laid out.” He sees the datacenter and robotics as the two largest markets available to Qualcomm, which “adds a steeper slope to our growth opportunity. . . . I think, over time, investors will see it and will get rewarded for it.”

There are countless execution steps between here and there, but I can’t wait to see what unfolds.